FrustumFormer: Adaptive Instance-aware Resampling for Multi-view 3D Detection.
Yuqi WangYuntao ChenZhaoxiang ZhangPublished in: CoRR (2023)
Keyphrases
- multi view
- single view
- multiple views
- three dimensional
- camera calibration
- depth map
- d objects
- dynamic scenes
- multiple cameras
- semi supervised
- range images
- view synthesis
- multi view learning
- detection method
- scene reconstruction
- surface reconstruction
- multi view images
- multi view reconstruction
- multiple viewpoints
- depth video
- multi view semi supervised learning
- multi view stereo
- visual hull
- co training
- pose estimation
- supervised learning
- object detection
- data points
- optical flow